ABSTRACT Advancements in pediatric drug development require innovative approaches that overcome challenges in assessing age-dependent drug disposition and toxic risks related to metabolism. Cytochromes P450 dominate drug metabolism yet roles for individual enzymes depend on genetics, disease states, co-medications, and ontogeny. Failure to account for those differences contributes to dosing challenges and possibly toxicity as reported for dextromethorphan, midazolam, and phenytoin. The NICHD Pediatric Formulation Initiative (PFI) workshops emphasized the need for better modeling to describe and predict how drug metabolism changes for children. Current pharmacokinetic models predict how drug clearance changes with age that affect the optimal dose, but those models are limited in two ways; (1) they require experimentally determined kinetic data for several enzymes that is not often available, and (2) they do not model formation of specific drug metabolites, which is important in predicting toxicity and drug interactions, regardless of whether clearance changes with age. We hypothesize that computational models of mixtures of P450 enzymes can predict how the ?metabolic fate?, i.e. the kinetics of drug metabolism and the resulting metabolite structures, of drugs changes with age. We propose building hierarchical mathematical models that at first predict the drug metabolites formed by metabolic enzymes (Aim 1), then the efficiency of formation for each metabolite (the kinetics) (Aim 2), and combine these models to predict metabolites formed by age-specific mixtures of P450s (Aim 3). This proposal makes significant steps toward achieving PFI goals. First, datasets created for this study will be made publicly available to foster model refinement and validation by the community. Second, simulation of metabolite profiles would yield tractable biomarkers and support studies on possible age-dependent drug-drug interactions, off- target biological activities, pro-drug activation, and formation of toxic species. Third, the models will indicate, for both new and existing drugs, when metabolic fate (consequently, toxicity and interactions) changes in pediatric patients, even when pharmacokinetics stay the same. Fourth, though not the primary aim of this study, successfully modeling metabolic efficiency could enable pharmacokinetic studies for predicting problematic pediatric drugs prior to carrying out any necessary experimental kinetic studies. Taken together, this proposal lays a strong foundation for developing models relevant that resolve challenges in optimizing drug dosages and minimizing toxicity risks for children.